Nasim435/Qwen-3B-Automotive-4000

QWEN-3B MODEL CARD / AUTOMOTIVE DOMAIN ■ ONLINE QWEN-3B
AUTOMOTIVE BASE → Qwen/Qwen2.5-3B-Instruct  ·  METHOD → QLoRA / 4-bit NF4  ·  SAMPLES → 4,000 QLoRA Automotive PEFT SFT 4-bit NF4 bfloat16 TRL BitsAndBytes Apache 2.0 01 OVERVIEW

Domain-adapted variant of Qwen2.5-3B-Instruct, fine-tuned on automotive instruction-following data using QLoRA (4-bit quantized LoRA fine-tuning). Created as an experimental learning and research project focused on modern parameter-efficient fine-tuning workflows in 2026-style LLM engineering.

Specialized toward automotive-related question answering and technical explanations. Trained on a curated subset of 4,000 samples from the BAAI automotive industry instruction dataset.

Model Identity BASEQwen/Qwen2.5-3B-Instruct DEVELOPED BYAlibaba Cloud / Qwen Team FINE-TUNED BYNasim Raj Laskar LANGUAGEEnglish LICENSEApache 2.0 Capabilities ► Automotive Q&A ► Diagnostic Explanations ► Vehicle Systems Knowledge ► Maintenance & Repair Instructions ► PEFT / QLoRA Experimentation 02 TRAINING DATA

Fine-tuned on the English subset of BAAI/IndustryInstruction_Automobiles. A custom subset of approximately 4,000 automotive instruction samples was selected for training.

Dataset Processing Pipeline FORMATQwen chat-template conversations STRUCTUREsystem → user → assistant SYSTEM PROMPT"You are an automotive expert assistant." SAMPLES~4,000 curated instruction pairs ◆ DATASET CONTENTS

The dataset consists primarily of automotive technical Q&A, diagnostic explanations, vehicle systems knowledge, and maintenance and repair related instructions.

03 TRAINING CONFIGURATION MethodQLoRA Quantization4-bit NF4 Precisionbfloat16 OptimizerAdamW 8bit Max Seq Len512 PackingDisabled Infrastructure PLATFORMAWS SageMaker INSTANCEml.g5.xlarge GPUNVIDIA A10G (24GB VRAM) Hyperparameters EPOCHS1 LEARNING RATE2e-5 BATCH SIZE4 (grad accum × 2) ► LoRA Adapter Parameters Rank (r)16 Alpha (α)32 Dropout0.05 Trainable Params~30M Total Params~3.1B Libraries Transformers · TRL · PEFT · BitsAndBytes Accelerate · Datasets Fine-Tuning Stack QLoRA · PEFT (LoRA adapters) SFT · 4-bit NF4 quantization 04 INTENDED USE ► Automotive Question Answering ► Educational Demonstrations ► PEFT / QLoRA Experimentation ► Automotive Technical Assistance ► LLM Fine-Tuning Research ► Domain Adaptation Studies 05 EXAMPLE USAGE Python

# Load model from Hugging Face Hub
from transformers import AutoTokenizer, AutoModelForCausalLM
model_name = "Nasim435/Qwen-3B-Automotive-4000"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model     = AutoModelForCausalLM.from_pretrained(
    model_name,
    device_map="auto"
)
prompt   = "Explain symptoms of a failing alternator."
messages = [{"role": "user", "content": prompt}]
text = tokenizer.apply_chat_template(
    messages,
    tokenize=False,
    add_generation_prompt=True
)
inputs  = tokenizer(text, return_tensors="pt").to(model.device)
outputs = model.generate(
    **inputs,
    max_new_tokens=150,
    temperature=0.7,
    top_p=0.9,
    do_sample=True
)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

06 LIMITATIONS ⚠ WARNING

• Experimental fine-tuned model — not intended for production safety systems
• May hallucinate or generate inaccurate automotive advice
• Not suitable for safety-critical or professional mechanical decision-making
• Trained on a relatively small subset (~4k samples); generalization may be limited

07 ACKNOWLEDGEMENTS Qwen TeamAlibaba Cloud BAAIDataset Contributors Hugging FaceEcosystem TRL / PEFTContributors BitsAndBytesQuantization AccelerateFramework DatasetsHuggingFace Open-SourceLLM Community NASIM RAJ LASKAR Qwen-3B-Automotive-4000
Apache 2.0 · English · 2026